Dataset Versions
Versions
v6
2024-09-28 8:08am
Generated on Sep 28, 2024
1337 Total Images
View All Images
Dataset Split
Train Set 89%
1185Images
Valid Set 8%
104Images
Test Set 4%
48Images
Preprocessing
Auto-Orient: Applied
Resize: Stretch to 640x640
Augmentations
Outputs per training example: 3
Flip: Horizontal, Vertical
90° Rotate: Clockwise, Counter-Clockwise, Upside Down
Crop: 5% Minimum Zoom, 18% Maximum Zoom
Rotation: Between -45° and +45°
Shear: ±37° Horizontal, ±29° Vertical
Grayscale: Apply to 100% of images
Hue: Between -58° and +58°
Saturation: Between -83% and +83%
Brightness: Between -53% and +53%
Exposure: Between -27% and +27%
Blur: Up to 1.6px
Noise: Up to 4.62% of pixels
Cutout: 25 boxes with 9% size each
Mosaic: Applied
Bounding Box: Flip: Vertical
Bounding Box: 90° Rotate: Clockwise, Upside Down
Bounding Box: Crop: 22% Minimum Zoom, 61% Maximum Zoom
Bounding Box: Rotation: Between -35° and +35°
Bounding Box: Shear: ±16° Horizontal, ±22° Vertical
Bounding Box: Brightness: Between -40% and +40%
Bounding Box: Exposure: Between -41% and +41%
Bounding Box: Blur: Up to 10.2px
Bounding Box: Noise: Up to 0.81% of pixels